Jhanvi Nagori

I've owned the backend for VoiceToNotes, a production SaaS product serving 150K+ users. I rebuilt its v3 architecture and reduced Firebase reads from 1M to 40K per day, cutting infrastructure cost and removing read-quota risk.

I shipped subscription billing through Razorpay and RevenueCat, resolved international recurring-payment failures, and built real-time note sharing, publishing workflows, and an FCM-based notification engine. I've also implemented GDPR/HIPAA-aligned data handling and scoped SSL pinning and TLS 1.3 for regulated markets.

At Web3Task, I independently built the frontend and backend of a Chrome extension for store review within eight weeks. I also delivered authentication, verified ad integrations, subscription syncing, and acquisition pages that raised Lighthouse scores from 20 to 100.

Alongside production work, I build AI systems end-to-end: AegisFlow, an LLM compliance gateway with PII masking and Kafka audit logging, and MediaStance Analytics, a real-time NLP classifier that has served 38,000+ predictions. I'm completing a Computer Science degree focused on AI/ML and looking to build backend or AI/ML products with product-driven teams.